mirror of
https://github.com/HKUDS/nanobot.git
synced 2026-08-09 13:58:36 +03:00
95160a304d
Adds two opt-in agent tools for controlling a computer: - computer_use: pixel-based (screenshot + mouse/keyboard) via a desktop (pyautogui) or browser (playwright) backend. - browser: DOM/accessibility-based web automation (act by element ref), reliable across ANY tool-calling model, not just vision/CU-trained ones. Core enabler in the runner: a tool may return image content blocks, which are split out and delivered to the model as a follow-up user message (_split_tool_result_media), so screenshots reach any vision provider (e.g. via OpenRouter/openai-compat) without provider-specific code. Both tools are OFF by default (tools.computerUse.enable / tools.browser.enable), are not exposed to subagents, and the browser tool supports an allowed_domains allowlist. Heavy deps (pyautogui/pillow/playwright) are an optional [computer-use] extra. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
25 lines
750 B
HTML
25 lines
750 B
HTML
<!doctype html>
|
|
<html><head><meta charset="utf-8"><title>Login</title></head>
|
|
<body>
|
|
<h1>Sign in</h1>
|
|
<form onsubmit="return go(event)">
|
|
<p><input id="user" aria-label="Username" placeholder="Username"></p>
|
|
<p><input id="pass" type="password" aria-label="Password" placeholder="Password"></p>
|
|
<button type="submit">Log in</button>
|
|
</form>
|
|
<div id="err"></div>
|
|
<script>
|
|
function go(e){
|
|
e.preventDefault();
|
|
var u = document.getElementById('user').value;
|
|
var p = document.getElementById('pass').value;
|
|
if (u === 'admin' && p === 'secret') {
|
|
location.href = 'dashboard.html?user=' + encodeURIComponent(u);
|
|
} else {
|
|
document.getElementById('err').textContent = 'Invalid credentials';
|
|
}
|
|
return false;
|
|
}
|
|
</script>
|
|
</body></html>
|